©Author(s) (or their employer(s)) 2026. No commercial re-use. See Permissions. Published by Baishideng Publishing Group Inc.
Explainable electroencephalography-based attention-deficit/hyperactivity disorder detection model with a combination of ternary pattern and twin wavelet transform
Yavuz Atas, Serkan Kırık, Kübra Yıldırım, Burak Tasci, Prabal Datta Barua, Ferhat Balgetir, Sengul Dogan, Turker Tuncer, Ru-San Tan, Elizabeth Palmer, Aruna Devi, U Rajendra Acharya
Yavuz Atas, Department of Pediatrics Division of Pediatric Neurology, Van Research and Education Hospital, Health Sciences University, Van 65030, Türkiye
Serkan Kırık, Department of Pediatrics Division of Pediatric Neurology, Elazig Fethi Sekin City Hospital, Elazığ 23100, Türkiye
Kübra Yıldırım, Sengul Dogan, Turker Tuncer, Department of Digital Forensics Engineering, College of Technology, Firat University, Elazığ 23119, Türkiye
Burak Tasci, Vocational School of Technical Sciences, Firat University, Elazığ 23119, Türkiye
Prabal Datta Barua, School of Business (Information Systems), University of Southern Queensland, Australia, Quilpie 41142, Queensland, Australia
Ferhat Balgetir, Department of Neurology, Firat University, Elazığ 23119, Türkiye
Ru-San Tan, Cardiology, National Heart Centre Singapore, Singapore 169609, Singapore
Elizabeth Palmer, School of Women’s and Children’s Health, University of New South Wales, Randwick 2031, Australia
Aruna Devi, School of Mathematics, Physics and Computing, University of Southern Queensland, Springfield 2031, Australia
U Rajendra Acharya, School of Mathematics, University of Southern Queensland, Springfield 2031, Australia
Co-corresponding authors: Burak Tasci and Sengul Dogan.
Author contributions: Tasci B and Dogan S contributed equally as corresponding authors. Both provided overall supervision of the study, coordinated collaboration among multidisciplinary contributors, and ensured methodological consistency, analytical accuracy, and intellectual integrity of the work. They jointly contributed to the interpretation of findings, critical revision, and final approval of the manuscript. Atas Y, Kırık S, Yıldırım K, Tasci B, Dogan S, and Tuncer T contributed to methodology; Dogan S and Tuncer T contributed to software; Atas Y, Yıldırım K, Tasci B, Barua PD, and Acharya UR contributed to validation; Atas Y, Kırık S, Yıldırım K, Tasci B, Barua PD, and Dogan S contributed to formal analysis; Atas Y, Kırık S, Yıldırım K, and Dogan S contributed to investigation; Atas Y, Kırık S, Balgetir F, and Acharya UR contributed to resources; Yıldırım K, Dogan S, and Tuncer T contributed to data curation; Atas Y, Kırık S, Yıldırım K, Tasci B, Barua PD, and Dogan S contributed to writing - original draft preparation; Tasci B, Barua PD, Balgetir F, Dogan S, Tuncer T, Tan RS, Palmer E, Devi A, and Acharya UR contributed to writing - review and editing; Yıldırım K, Dogan S, and Tuncer T contributed to visualization; Tan RS, Palmer E, and Acharya UR contributed to supervision; Acharya UR contributed to project administration. All authors contributed to conceptualization and have read and agreed to the published version of the manuscript.
Institutional review board statement: The study was approved by the Ethics Committee of Firat University, No. 2023/02-02.
Informed consent statement: Informed consent was obtained from all subjects involved in the study.
Conflict-of-interest statement: The authors report no relevant conflicts of interest for this article.
Data sharing statement: The data presented in this study are available on request from the corresponding author. The data are not publicly available due to restrictions regarding the ethical committee institution.
Corresponding author: Sengul Dogan, Full Professor, Department of Digital Forensics Engineering, College of Technology, Firat University, Elazığ 23119, Türkiye.
sdogan@firat.edu.tr
Received: August 11, 2025
Revised: September 9, 2025
Accepted: November 11, 2025
Published online: March 19, 2026
Processing time: 201 Days and 0.9 Hours
BACKGROUND
Attention-deficit/hyperactivity disorder (ADHD) is a common neurodevelopmental condition characterized by inattention, impulsivity, and hyperactivity. Traditional diagnosis relies on clinical evaluation, which is time-consuming and subjective. Electroencephalography (EEG) signals provide an objective alternative, and machine learning methods can improve their diagnostic utility.
AIM
To develop an explainable EEG-based model for ADHD detection by integrating a novel combination ternary pattern (CTP) feature extractor with twin wavelet transform (TWT) for multilevel signal analysis, and to evaluate its effectiveness in providing accurate, channel-wise, and fusion-based classification results for objective and rapid ADHD diagnosis.
METHODS
A new EEG dataset containing more than 7000 segments from 137 ADHD patients and 150 controls was studied. A novel feature engineering framework was developed, combining a new CTP extractor with statistical features. A multilevel feature extraction structure was designed using a newly proposed TWT for signal decomposition. Extracted features were reduced to the most informative 263 using neighborhood component analysis. Channel-wise classification was performed with k-nearest neighbors, followed by iterative majority voting across 20 EEG channels.
RESULTS
Single-channel analysis achieved up to 99.12% accuracy. By applying majority voting, overall classification accuracy increased to 99.97%, with similarly high sensitivity and specificity.
CONCLUSION
Our study introduces a large ADHD EEG dataset and a novel model integrating TWT and CTP. The model provides highly accurate, channel-wise, and fusion-based results, offering a promising objective tool for rapid ADHD diagnosis.
Core Tip: This study introduces a novel combination ternary pattern-based framework integrated with a newly developed twin wavelet transform for automated attention-deficit/hyperactivity disorder detection using electroencephalography signals. Leveraging a newly acquired multi-channel electroencephalography dataset of over 7000 recordings, the proposed approach performs channel-wise feature extraction, statistical fusion, and optimal feature selection via neighborhood component analysis. The model achieves remarkable classification performance, with up to 99.97% accuracy through majority voting, demonstrating its potential as a reliable, explainable, and non-invasive diagnostic support tool for attention-deficit/hyperactivity disorder detection assessment.
